Adaptive image contrast enhancement based on human visual properties

摘要: Existing methods for image contrast enhancement focus mainly on the properties of to be processed while excluding any consideration observer characteristics. In several applications, particularly in medical imaging area, effective diagnostic purposes can achieved by including certain basic human visual properties. Here authors present a novel adaptive algorithm that tailors required amount based local and observer's Just-Noticeable-Difference (JND). This always produces adequate output image, results almost no ringing artifacts even around sharp transition regions, which is often seen images conventional techniques. By separating smooth detail areas an considering dependence noise visibility spatial activity treats them differently thus avoids excessive noise, another common problem many existing The JND-Guided Adaptive Contrast Enhancement (JGACE) technique very general applied variety images. particular, it offers considerable benefits digital radiography applications where objective increase utility A detailed performance evaluation together with comparison techniques given demonstrate strong features JGACE. >